a SoR project for all

[Private Reid]

Err, my cousin is doing someething to do with robotics at his UNI

Something about colour detection but in a really dense level, ill ask if he would like to say something.

[benji]

ive been checkin this thread everynow and then and apparently no one is donating to write articles,,
the thing is that i dont wanna waste time to make mine while the book would still be poor of articles,,

i would consider starting when there are more than 12 articles,,, lets wait n see

[Private Reid]

Well my cousin isn't doing exactly what i thought, it's more to do with a computer realising what it's looking at.

And sorry i can't help with the articles, I'm not the brightest of people...

[Kohanbash]

airman00-
Hey I have my chapter done. Where should I submit it to?
Thanks,
sdk32285

[airman00]

nice!


post it up in the members tutorials place for review

[airman00]

dudes we need more volunteers!!!!

If we don't get enough algorithms chapters then I might have to include member tutorials and some of Admin's tutorials into the book.

[Webbot]

dudes we need more volunteers!!!!

If we don't get enough algorithms chapters then I might have to include member tutorials and some of Admin's tutorials into the book.

Hmmm. So you 'sell' a book of 'our' freely donated tutorials? So: as a previous tutorial contributor do I get paid if my tutorial is included - and do you need my permission? As a potential purchaser of the book - do I get any more info over, and above,  what is freely available on the SoR website?

Not wanting to be negative but: unless it contributes lots of new stuff then it sounds like a directory of websites.

On a positive note - I like money! So if there are any tutorials 'missing' then I may be interested in contributing. Let me know subjects, deadlines, rewards etc.

Just to remind/motivate me: can I have the 'elevator pitch' - ie you have 15 seconds, in between floors,  to tell me what your book does, what I will learn, why I should buy it, why should I invest time/money in helping to create it, who is it targeted at !  Need to make sure its not a collection of rambling, un-connected, thought processes. Are you planning to sell via outlets: Borders, Amazon etc

Hey airman00: dont get me wrong - I dig 'the idea' - but think it currently lacks focus.

[airman00]

I understand your concern

The plan is as follows: Collect a lot of tutorials of algorithms

Make a free eBook of it

Then after that is all done I will discuss with the other authors how the money issue will work
I promise I will not sell anything without the author's permission

It is basically a compilation of tutorials that available online. Its easier for people to buy a book with all the stuff organized and relevant to each other , than to search online for tutorials.  This book is aimed for people who perhaps do not have access to computers or would like a book for reference. People who have never heard of SoR can stumble onto this book on Amazon or something. I personally prefer a hard copy then an online copy.

Sure elevator pitch it is.

[benji]

my plan was to pick somthing (like vision or voice recognition) and everybody in SoR contribute in the research
so everyone can donate and benefit

id rather get back to my first idea than to continue in this book ,,as long as no one would volanteer to write somthin valueble

[Webbot]

Thanks airman00

Ok - so that all kinda makes sense. I agree that a printed version is a lot easier to flick through and get ideas - unlike web searching where you need to know what it is that you are looking for before you start. And an eBook means you have no print costs.

If it also helps to promote SoR, and in turn generate more members with more ideas; more tutorials etc then this is a good thing for the robo-world and all of us.

I still believe that it would be useful to have a high-level view of the layout (ie chapter by chapter) of the book ie what are you covering e.g. micro-controllers, sensors, servos, motors etc and are you covering electronics and/or software? If its both then can a newbie get the book and then build something from start to finish - if so 'what'? You say 'a collection of tutorials and algorithms' - which sort of implies 'both'. Equally you say its for 'people who perhaps do not have access to computers' - so it cant just be algorithms. More importantly: a 'chapter by chapter' layout would identify what you need and then what's currently missing from the tutorials. Without this I still think you will have a book that is an un-focused collection of 'stuff' - and if someone buys it they may be surprised to realize that there isn't enough info to actually build a complete 'thing' (other than the $50 robot - say). Equally: would a more experienced person buy a 200 page book if the only bit they were interested in were pages 90 to 95 ? So: its almost like a project plan - what do you need, by when, and who will produce the missing stuff. I would suggest that on-going Forum posts aren't the place to have this plan as it becomes too long and rambling to try and find out what is the current status. It would be easier to have one page that you can continuously keep up to date. But, hey, thats just my view and I wish you well !!

Personally: I am happy for you to use my Stepper Motor and Motor Driver tutorials if they are relevant and good enough.
Commercially: I wouldn't say 'no' to a small payment for the time I have spent producing them (and to further fund my hobby!) - but beyond that I would gladly donate the money to a charity or to help fund any other SoR project/competition. (Seems like sending cheap components to Australia would be a big hit with the people there  )

[airman00]

OK guys - heres what I think :

The book will be only algorithms. Some complex , some simple . My vision for this book is a book filled with algorithms that are explained very clearly.
Here are the categories of algorithms I want to include:
        1.  I want it to have algorithms for the types of sensors out there - sonar , IR , accelerometer, etc. A lot of information will be taken from the Forum about how to interface all those sensors. Only the common sensors of each sensor type will be selected - e.g. For sonar we will use the PING, the EZ1, and the SRF05. Each sensor would have around a page or two describing how it works and the code for it in BASIC and C .
       2.  Then there will be more algorithms that are not specifically for a certain sensor , but rather are generic - like maybe a A2D reading to matrix.
       3.   The other category of algorithms for the book would be algorithms for filtering ,such as the Kalman Filter.
       4.  If enough information is collected we can also do algorithms relating specifically to vision.
       5.  The final category would be algorithms relating specifically to types of robots. I understand that there are many ways to build a biped, so it will be the preference of the author of the biped tutorial to choose the design . Some types of robots I'm thinking of - sumo, line follower, biped, spider, quadraped,robot arm,follower robot.

The book will be a pretty big book and a majority of information will be taken from the forum - so that means less articles need to be written from scratch. I am not aiming to include every single sensor out there , but I do aim to include a lot of sensor-specific algorithms.

What do you guys think?

Also, I guess I'll have to extend the deadline for articles to another 6 months from now from lack of articles. If anyone has written one up please post it up by a member tutorial for review by the community.

[Admin]

I think this sounds like a good idea. So, its going to be a big printable pdf file? Any pseudocode or actual code examples in it?

I can write a preface for the book, and if I have time I'll maybe include a new algorithm I've invented (but haven't made public yet).

And just a reminder, any algorithm tutorials people write can also be entered into my bi-annual robot contest.

[airman00]

It will at first be a pdf file and then we can publish it into a book ( work out the money issues later)

It will contain both pseudo code and working code in BASIC and C. Code will be written in a way that will work for most microcontrollers with little modification. Meaning it will not be microcontroller specific. So instead of the port name PortA.1 you use ServoPort1.

[Webbot]

Ok so here are a few more thoughts:-

1. Sensors - great idea C/Basic. So if someone supplies a C tutorial for a given sensor then you will need someone to translate into Basic (and test it) and vice-versa.

2. Actuators. I think they will be harder then sensors. With a sensor the manufacturer dictates how they work. With a motor (non-servo) then you need some kind of controller and they all work differently. So: if you are avoiding the electronics side of things then do you plan to show algorithms for a finite number of different motor controllers? For example: Pololu has a DC motor controller driven purely via UART where as (most) other solutions are H-Bridge linked and have different numbers of wires and ways to control speed, braking, direction etc. Outputs are obviously important - as there is no point in monitoring sensors if there is no output !? Maybe that is why your overview doesn't mention them?

3. Print. Am not sure what experience you have with professional printing and the complications of say Pantone spot colours, cmyk vs rgb, image resolutions, bleed etc - let alone 'design'. Most people fail to realise that professional printing is a completely different world to, say, printing something at home/work or for the web and is much more complex. If you can satisfy the needs of a professional printing company then its normally easy to produce a web PDF but going the other way normally requires a vast amount of re-work (at least getting higher resolution image of every image!). If you supply a Word file to the printer then they will normally charge you to redo each page in a DTP system as desktop word processors just cannot do 'print'. I say this from a position of knowledge as that is what the commercial software I write for my own company actually does!! I suggest it may be worth you looking at something like Apache FOP (coz its free) - if you know Java. The basics are that the content text is stored in an XML format with mark up like 'Main Heading', 'Sub Heading', 'Bullet Point' etc etc. You then have an XSL template that decides what font, point size etc etc is used for each of these mark up styles. The template can feed into FOP and produce a PDF. So if you suddenly decide you want page numbers or make all the section headings slightly bigger, then just change the template, and create a new PDF. You can also create content/index sections with page number references.  But you also need to manage images - ie the image you put into a downloadable PDF may be 72dpi but, for a printer, you need to have a minimum of say 300dpi. So you need some imaging software to create the different resolutions for you. Our software would be an ideal solution as everyone could author stuff concurrently over the web and then we can produce web PDFs as well as print quality PDFs from the same data set - however I think it could require more time to set up and train than I can currently afford.

4. Either way you will need to enforce some rules re images. If you are only focusing on algorithms then (luckily) you wont have many photos - but you may have flow charts. So what format should these be in: ie rasterized rgb (jpg/png) or vector based formats like svg. Only suggesting this as you may otherwise end up with a whole bunch of image formats which someone then has to pull together and sort out.

5. If you really are contemplating having the book printed (and you aren't familiar with some of the points mentioned above) then I would strongly urge you to send something to your local friendly printing company quite early on to find out their view as to where they would need to spend any time with their own internal graphic designers so that you can see if you can fix this at source yourself, and not have to pay the printing company to: lay out the whole book again for you, redo all the images, etc at vast expense.

[airman00]

Thanks Webbot for those thoughts - they really helped

I'm writing up an outline based on your ideas and mine.

Right now I am concerned with gathering all the material . You bring a great point about contacting the printing company. My plan is to just keep on gathering information and then after I have it all, I will arrange it on a page per request of the printing company.

I'll make a new post requesting articles ( which could be 1/2 page long even).

Thanks for the assistance Webbot .

[airman00]

http://www.societyofrobots.com/member_tutorials/node/115

I updated it

I really need some volunteers who can help write articles .
Even newer people to robotics can contribute. Simply scavenge the forum for information on how interface a part ( lets say a PING sensor ) and then write a short comprehensive tutorial that can show someone how to use that sensor. Please follow the guideline on the Book Project page. Authors names will be mentioned in the book next to their respective articles.

I'll update the Book Project page again later tonight. For the record I will be doing the CMUcam, bipedal robot,and whatever else is available.

If people can submit to me the articles they are working on that would be great  , so that way I won't end up writing a tutorial on something that someone is already doing.
If you are using some sensor anyway on your robot , can you please document it so that it can be included into this book?

Thanks,
Eric

[rootoftwo]

Hey, I'm new here. I'm not sure if this topic is still alive - but if it is, you should consider using http://www.lulu.com/ - a web-based print-on-demand service. There are a multitude of formats and templates with decent instructions on how to get what you want. There are even options for distribution through bookstores. I've used it for an exhibition catalog I edited and designed http://stores.lulu.com/rootoftwo - I've made it available as a free download (80MB - its at print resolution) and have printed several hundred copies (at manufacturing cost - since I am the creator and publisher). For books sold through online bookstores Lulu take a 25% commission of the creator revenue (or 19¢, whichever is greater) that you, the creator, set for the items you publish. For example, if set your creator revenue to $4.00 (on top of the cost of printing and binding the book) Lulu add a $1.00 commission. N.B. I got funding for this project through a grant and never needed it to be a commercial success. I have been able to give everyone involved in the project a stack of copies as well as getting it into school libraries ($99 for an ISBN number). This might be a low-risk strategy for what you are talking about here. I hope this is useful. (BTW I don't work for Lulu - also consider http://www.blurb.com/)

Cheers,
John.

[airman00]

Sadly I think this project is dead .... 
Maybe we can revive the project in a while ....

[frodo]

oh, thats a pity i had some ideas. oh well.